High-throughput full-field thickness mapping of semiconductor wafers via hyperspectral imaging with a field-modulated...

Area of Science:

• Materials Science
• Optical Metrology
• Industrial Process Control

Background:

• Thin-film thickness is critical for material properties.
• Point-scanning methods are inadequate for large-scale, real-time uniformity analysis.
• Existing methods struggle with spectral distortions in large-field imaging.

Purpose of the Study:

• To develop a high-throughput, full-field online method for thin-film thickness monitoring.
• To address and resolve the spectral drift phenomenon in hyperspectral imaging (HSI).
• To enable precise characterization of large-area film uniformity in industrial settings.

Main Methods:

• Designed a high-uniformity linear illumination system for HSI.
• Developed a physical model coupling specular reflection and diffuse scattering.
• Introduced an effective interface roughness coefficient to correct for substrate scattering.
• Utilized a least-squares nonlinear fitting algorithm (LSNFA) for thickness extraction.

Main Results:

• Successfully decoupled substrate-induced diffuse scattering, resolving spectral drift.
• Achieved accurate thin-film thickness extraction with spectral baseline correction.
• Demonstrated full-size (4-inch wafer) thickness mapping in under 60 seconds.
• Obtained an absolute error of less than 1 nm compared to ellipsometry.

Conclusions:

• The proposed HSI method provides a precise and efficient solution for industrial thin-film thickness metrology.
• This technique overcomes limitations of traditional methods for large-area uniformity assessment.
• The study offers a robust approach for real-time quality control in mass production.